366 Kees Hengeveld
- A model of encoded intentions and conceptualizations
Since the introduction of the layered structure of the clause in FG
(Hengeveld 1988, 1989; Dik 1997), the question has been raised (e.g. by
Bolkestein 1992 and Harder 1996) whether this structure should be inter-
preted as a representation of the communicative process itself, or as a
representation of the linguistic units put to use in that process. These two
interpretations have become known as the process and pattern views on
underlying representations. In this volume Fortescue raises the same issue
explicitly in relation to FDG, and concludes that it is a process model. An-
stey, Bakker and Siewierska, Harder, and Nuyts all arrive at the same
conclusion.
It is important to note that this is not how I intend FDG to be taken. I
conceive of FDG as a pattern model, i.e. as a model that represents linguis-
tic facts. The process interpretation given to it by the aforementioned
authors is probably due to two facts. First of all, FDG is presented in
Hengeveld (this volume) as the grammatical component of a wider theory
of verbal interaction (cf. Dik 1997) and as such interacts with a conceptual
and a contextual component. These latter components, however, are not
part of the grammar as such, even though many grammatical phenomena
can be studied much more fruitfully if such components are assumed to ex-
ist. Secondly, in FDG, as in FG, the patterns of language are described as
reflecting the process of communication. This, however, does not mean to
say that FDG is a model of that process.
The latter point can be clarified starting from a familiar and generally
accepted underlying representation in FG:
(1) (e 1 : [(f 1 : collapseV (f 1 )) (x 1 : houseN (x 1 )Ø)Pat] (e 1 ))
‘the collapsing of a house’
The representation in (1) contains three types of variable from the re-
presentational level: (e) for states-of-affairs, (f) for relations and properties,
and (x) for spatial objects. These variables are defined in terms of entity
types in the external, extra-linguistic world. Yet nobody would claim that
the formula in (1) represents part of the external world, as it would if it
were an ontological representation. Rather, (1) represents a linguistic unit
in terms of its function of designating part of the external world. This fact
has been obscured by informal terminological usage. It is common to say
that (1) represents a state of affairs, whereas it would be more appropriate
to say that (1) represents a linguistic unit in terms of its ideational function.